ur-agent 1.79.1 → 1.80.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/CHANGELOG.md CHANGED
@@ -1,5 +1,55 @@
1
1
  # Changelog
2
2
 
3
+ ## 1.80.1
4
+
5
+ - Removed the complete unreachable `ultraplan` implementation after its public
6
+ command was disabled: stale task types, pills, metadata, polling, prompt
7
+ routing, and UI state can no longer create a half-supported workflow. The
8
+ active `ultrareview` workflow remains supported and unchanged.
9
+ - Renamed the `/fix-bug` source module and test from their obsolete `debug-v2`
10
+ filenames, without changing the public command contract.
11
+ - Recorded the retracted Agent-to-Agent version finding: UR deliberately serves
12
+ truthful legacy 0.3 and modern v1 cards through negotiation, so no protocol
13
+ behavior was changed.
14
+ - “Approve all”/`autoApprove` remains supported and unchanged.
15
+
16
+ ## 1.80.0
17
+
18
+ - Sandbox security now supports trusted-source-only credential masking and
19
+ injection, regex/JWT masking, AWS credential pairs, SigV4 rewriting,
20
+ TLS-terminating egress, denied domains, and a strict network allowlist.
21
+ Trailing-slash deny paths are regression-tested. Project settings cannot
22
+ enable credential injection or weaken these boundaries.
23
+ - Permission rules now match tool input parameters with wildcards, including
24
+ `Agent(model:opus)` and `Bash(timeout:*)`. Bash and PowerShell approval checks
25
+ detect commands hidden in zsh conditionals, display tabs/control/zero-width/
26
+ bidirectional characters visibly, and reject quoted-path ambiguity.
27
+ - Screen-reader mode adds append-only plain-text output, edit announcements,
28
+ reduced animation, `--screen-reader`, `UR_SCREEN_READER`, and
29
+ `/config screenReader=true`. Vim users can set an insert-mode escape sequence
30
+ such as `/config editor=vim vimEscape=jj`.
31
+ - Model Context Protocol tools can return final `input_required` continuations;
32
+ older elicitation remains compatible. Roots now include authorized added
33
+ directories and emit change notifications. The stateless Tasks/Apps command
34
+ is now clearly named `ur mcp serve-web`.
35
+ - WebSearch has an atomic per-session budget (200 by default), subagent spawn
36
+ totals warn without breaking long runs, and `--forward-subagent-text` exposes
37
+ nested agent answers in stream JSON with parent tool-use correlation.
38
+ - OpenTelemetry adds privacy-safe assistant-response events, message/request/
39
+ tool correlation, and workflow name/run identity without enabling content
40
+ capture.
41
+ - `/config key=value`, the `DirectoryAdded` hook, `/session
42
+ status|list|archive|unarchive`, and matching `ur session` commands add direct
43
+ automation and conversation lifecycle control. Archived conversations remain
44
+ unavailable to resume/fork until explicitly restored.
45
+ - Public command cleanup renamed `/debug-v2` to `/fix-bug`, removed the no-op
46
+ `/output-style` and dormant `/ultraplan` registrations, removed redundant 3D
47
+ aliases, expanded protocol help text, and added registry assertions against
48
+ version jargon. “Approve all”/`autoApprove` remains supported and unchanged.
49
+ - Research, coding-impact analysis, Blender/OpenSCAD/3ds Max/custom-app 3D
50
+ pipelines, and the self-healing task-output fix from 1.79.x remain fully
51
+ documented and covered by the release suite.
52
+
3
53
  ## 1.79.1
4
54
 
5
55
  - Bash execution now survives external cleanup of its session task-output
package/README.md CHANGED
@@ -438,16 +438,16 @@ as first-class subcommands in the shipped CLI.
438
438
  | `ur auth gemini` | Use the official Gemini CLI login flow where supported. |
439
439
  | `ur auth antigravity` | Use the official Antigravity CLI login flow where supported. |
440
440
  | `ur config set` | Persist safe non-secret provider settings such as provider, model, base URL, command path, and fallback. |
441
- | `ur mcp` | Configure MCP servers, expose fail-closed built-in tools over stdio, or run the opt-in stateless MCP 2026 HTTP adapter with Tasks and Apps. |
441
+ | `ur mcp` | Configure Model Context Protocol servers, expose fail-closed built-in tools over standard input/output, or run the secure stateless web server with Tasks and Apps. |
442
442
  | `ur skill` | Initialize, run, strictly verify, Ed25519-sign, and create trusted signing keys for portable skills. |
443
443
  | `ur ag-ui serve` | Start the secure AG-UI HTTP/SSE adapter for user-facing applications with truthful capability discovery. |
444
444
  | `ur plugin` | Install, update, enable, disable, and validate UR plugins that can add MCP tools, skills, templates, validators, language adapters, LSP servers, agents, hooks, output styles, and commands. |
445
445
  | `ur role-mode` | Install built-in Architect, Code, Debug, and Ask role modes. |
446
- | `ur acp` | Run official-SDK ACP v1 over stdio, or manage the separate UR HTTP JSON-RPC server. |
446
+ | `ur acp` | Connect editors through the Agent Client Protocol, or manage the separate UR HTTP JSON-RPC server. |
447
447
  | `ur exec` | Run one or more prompts in non-interactive mode with optional concurrency. |
448
448
  | `ur ide diff` | Capture editor-readable inline diff bundles. |
449
- | `ur a2a card` | Print the v0.3 Agent Card preview; the live server negotiates a separate strict v1 card. |
450
- | `ur a2a serve` | Start opt-in A2A v0.3 plus v1 JSON-RPC/HTTP+JSON bindings and the separate UR compatibility task API. |
449
+ | `ur a2a card` | Print Agent-to-Agent discovery metadata. |
450
+ | `ur a2a serve` | Start the negotiated Agent-to-Agent server with automatic client compatibility. |
451
451
  | `ur sdk` | Show programmatic headless usage and scaffold SDK examples. |
452
452
  | `ur trigger` | Parse a GitHub/Slack webhook payload and optionally launch a headless UR run. |
453
453
  | `ur agent-templates` | List or install reusable project agent templates. |
@@ -531,7 +531,7 @@ setup, supported features, and limitations.
531
531
  New slash skills run agentic work in isolated git worktrees and leave changes
532
532
  local for review. They ask before the final full verification suite and do not
533
533
  commit, push, or create PRs unless explicitly requested:
534
- `/debug-v2`, `/refactor`, `/paper-implementation`, `/benchmark`, `/security-review`, `/dockerize`, `/latex-paper`.
534
+ `/fix-bug`, `/refactor`, `/paper-implementation`, `/benchmark`, `/security-review`, `/dockerize`, `/latex-paper`.
535
535
  Install matching agent templates with `ur agent-templates install`.
536
536
 
537
537
  New built-in tools (exposed through MCP and the UR HTTP compatibility API): GitHub, API, Browser, Docker, TestRunner, Database. File-system and terminal tools are already built in (FileRead, FileEdit, FileWrite, Glob, Grep, Bash, PowerShell).
@@ -644,7 +644,7 @@ ur skill run security-review "src/auth.ts"
644
644
  ur skill verify security-review --require-trusted
645
645
  ur skill keygen release --out ~/.ur/keys/release.pem
646
646
  ur skill sign security-review --key ~/.ur/keys/release.pem --key-id release
647
- UR_MCP_HTTP_TOKEN='<secret>' ur mcp serve-http --port 8976
647
+ UR_MCP_HTTP_TOKEN='<secret>' ur mcp serve-web --port 8976
648
648
  ur artifacts capture-tests --command "bun test"
649
649
  ur artifacts serve --port 4180
650
650
  ur agent-task pr --create --dry-run